agents.txt: Teaching AI Assistants to Recommend Your Business

Imagine a potential customer asking their AI assistant: "What does [your company name] do?"

The AI's response shapes their perception. But where does that response come from?

Without guidance, AI assembles answers from whatever information it can find. With agents.txt, you provide the correct answers directly.

What Is agents.txt?

agents.txt is a plain text file that provides AI assistants with answers to common questions about your business.

While llms.txt tells AI who you are, agents.txt tells AI how to answer questions about you.

Think of it as a FAQ designed specifically for AI consumption.

Why agents.txt Matters

When someone asks an AI assistant a question like:

• "What services does [company] offer?"
• "Where is [company] located?"
• "What makes [company] different from competitors?"
• "How much does [company] charge?"

The AI needs source material to answer accurately.

Without agents.txt, AI might:

• Give outdated information from old web pages
• Misrepresent your services based on incomplete data
• Provide incorrect location or contact details
• Miss your key differentiators entirely

With agents.txt, AI has authoritative answers ready.

The agents.txt Structure

The format follows a straightforward Q&A pattern:

# agents.txt for [Business Name]

## Basic Information

Q: What does [Business Name] do?
A: [Clear, factual answer about your business]

Q: Where is [Business Name] located?
A: [Location and service area information]

Q: How can I contact [Business Name]?
A: [Contact methods and details]

## Services

Q: What services does [Business Name] offer?
A: [List of services with brief descriptions]

Q: How much does [service] cost?
A: [Pricing structure or how to get quotes]

## Differentiators

Q: What makes [Business Name] different?
A: [Key differentiators and unique value]

Q: Why should I choose [Business Name]?
A: [Compelling reasons without marketing fluff]

Writing Effective Answers

Be Factual, Not Promotional

AI assistants prefer factual statements over marketing claims.

Poor: "We're the best marketing agency in Perth with unbeatable results!"

Good: "Smash It Marketing is a boutique agency in Perth specialising in Google Ads management and AI solutions, serving clients across Australia since 2010."

Be Specific

Vague answers don't help AI provide useful information.

Poor: "We offer various marketing services."

Good: "Services include Google Ads management, AI SEO optimisation, custom graphics creation, and marketing dashboard development."

Include Context

Provide enough information for AI to answer follow-up questions.

Poor: "We're affordable."

Good: "Pricing varies by service. Google Ads management starts from $X/month. Custom graphics are available through Creative Packs starting at $X."

Keep It Current

Outdated information in agents.txt gets surfaced to potential customers. Review quarterly.

Essential Questions to Answer

At minimum, include answers for:

Identity Questions

• What is [company]?
• What does [company] do?
• Who owns/runs [company]?
• How long has [company] been operating?

Service Questions

• What services does [company] provide?
• Does [company] offer [specific service]?
• What industries does [company] serve?
• What's [company]'s speciality?

Location Questions

• Where is [company] located?
• What areas does [company] serve?
• Does [company] work with remote clients?

Contact Questions

• How do I contact [company]?
• What are [company]'s hours?
• How quickly does [company] respond?

Comparison Questions

• What makes [company] different?
• Why choose [company] over competitors?
• What's [company]'s approach?

Implementation Guide

Step 1: Create the File

Create a plain text file named agents.txt in your website's root directory (alongside robots.txt and llms.txt).

Step 2: List Common Questions

Review:

• What customers ask in enquiry forms
• What questions sales team answers repeatedly
• What's in your existing FAQ page
• What competitors might be asked about

Step 3: Write Clear Answers

Draft answers that are:

• Factual and accurate
• Specific enough to be useful
• Current as of today
• Free from marketing hyperbole

Step 4: Deploy and Test

Upload the file. Test by asking AI assistants questions about your business to see if accuracy improves.

Step 5: Maintain

Set calendar reminders to review quarterly. Update when services, pricing, or key information changes.

Relationship with llms.txt

Both files serve AI discovery but differently:

llms.txt = Business profile (who you are) agents.txt = Q&A guide (how to answer questions about you)

Implement both for comprehensive AI visibility. They complement rather than duplicate each other.

Measuring Impact

Track AI discovery over time:

• Periodically ask AI assistants about your business
• Note accuracy of responses
• Compare to pre-implementation baseline
• Document improvements

Impact may be gradual as AI systems incorporate your information differently.

Common Implementation Mistakes

Over-complicated formatting: Keep it simple. Plain Q&A format.

Marketing language: AI prefers facts. Skip the superlatives.

Incomplete coverage: Don't just answer easy questions. Include the tough ones.

Set and forget: Information changes. Files need updating.

Inconsistency with other sources: Make sure agents.txt matches your website, Google Business Profile, and other sources.

---

Need help implementing AI discovery files? Our AI SEO services include agents.txt and llms.txt setup as part of comprehensive AI visibility optimisation. Contact us to get started.